Applied Mathematical Modelling 37 (2013) 384397

Contents lists available at SciVerse ScienceDirect

Applied Mathematical Modelling

journal homepage: www.elsevier.com/locate/apm

Modeling and dynamic simulation of mixed feed multi-effect evaporators

in paper industry
Deepak Kumar a,b,, Vivek Kumar b, V.P. Singh b
a
b

Department of Mathematics, Galgotias University, Greater Noida, UP, India

Department of Paper Technology, Indian Institute of Technology Roorkee, UK, India

a r t i c l e

i n f o

Article history:
Received 30 August 2010
Received in revised form 16 February 2012
Accepted 29 February 2012
Available online 8 March 2012
Keywords:
Multiple effect evaporator
Mixed feed
Fsolve
Ode45
Boiling point rise
Dynamic response

a b s t r a c t
A wide range of mathematical models for multiple effect evaporators in process industry
including paper industry are well reported in the literature but not so extensive work on
the dynamic behavior of MEE system is available in the literature. In the present study
dynamic behavior of multi-effect evaporator system of a paper industry is obtained by
disturbing the feed ow rate, feed concentration, live steam temperature and feed temperature. For this purpose an unsteady-state model for the Multi-effect evaporator system is
developed. Each effect in the process is represented by a number of variables which are
related by the energy and material balance equations for the feed, product and vapor ow.
A generalized mathematical model which could be applied to any number of effects and all
kinds of feeding arrangements like forward feed, backward feed, mixed feed and spilt feed
in the MEE system with simple modications was nally obtained. Finally model for mixed
feed sextuple effect falling lm evaporators system was solved using MATLAB. For the
steady state and dynamic simulation the fsolve and ode45 solvers in MATLAB source
code is used respectively.
2012 Elsevier Inc. All rights reserved.

1. Introduction
Pulp and paper industry is an energy intensive industry. Huge amount of raw material, chemical, energy and water are
consumed in the process of paper manufacturing. One such energy intensive subsystem, Multi-effect evaporator system
(MEE), is used to concentrate black liquor from pulp mill. Rao and Kumar [1] indicated that the Multi-effect evaporator
system alone consumes around 2430% of the total steam consumed in a large Indian paper mill.
The multiple effect evaporators (MEE) system in paper industry consists of a number of evaporators in series and large
number of permutation and combinations are possible for feeding of black liquor and live steam. Dynamic behavior of such
system can only be studied by mathematical modeling and simulation.
A wide range of mathematical models for multiple effect evaporators in process industry including paper industry are
well reported in the literature. These mathematical modeling of MEE system are based on linear and non-linear mass and
energy balance equations, and relationships pertaining to the physical properties of the solution. These mathematical models
are differentiated on the basis of the heuristic knowledge, which is incorporated in their development.
Various researchers have solved the mathematical models of MEE system of various process industries by using different
solution techniques to obtain steady state simulation [219]. Some of them also have performed steady state simulation for
MEE system of the paper industry [1,2,810,15,16,19,20] by solving the models using numerical techniques. The authors

Corresponding author at: Department of Mathematics, Galgotias University, Greater Noida, UP, India. Tel.: +91 8126421115; fax: +91 132 2714311.
E-mail address: dkr2009@gmail.com (D. Kumar).
0307-904X/$ - see front matter 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.apm.2012.02.039

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

385

Nomenclature
A
BPR
C
Cp
F
h
L
M
Pl
t
T
U
W
X
k

shell area, m2
boiling point rise, C
constant
specic heat of water at constant pressure, kJ/kg K
feed, kg/s
enthalpy, kJ/kg
liquor level, m
mass, kg
liquor density
time, s
temperature, C
overall heat transfer coefcient (OHTC), kW/m2 K
mass ow rate, kg/s
solid content, kg solids/kg total
latent heat of vaporization

Subscripts
c
condensate
i
effect number
l
liquor
v
vapor

[1,2,810,20] have used NewtonRaphson method with Jacobian matrix and method of Gauss elimination with partial pivoting supplemented with LU decomposition with the aid of Hilbert norms, and FORTRAN-77 language for computer programming. While the authors [15,16,19] simulated the MEE system using FORTRAN-90 and run the Pentium IV machine,
using Microsoft FORTRAN power station 4.0 compiler.
Although the methods of design and steady state operation of single and multi-effect evaporators are well documented in
the literature but not so extensive work on the dynamic behavior of MEE system is available in the literature. The dynamic
behavior of two or three effect MEE systems in process industries like sugar, food, desalination and paper etc. is studied by
few researchers [12,2126]. Miranda and Simpson [12] described a phenomenological, stationary and dynamic model of a
multiple effect evaporator (tomato concentrate) for simulation and control purposes. Andre and Ritter [21] adequately
described the open loop dynamic response of a laboratory scale double effect evaporator (for concentrating the sugar solution) by a simple mathematical model based on the unsteady state material and energy balances associated with the various
elements of the system. Bolmstedt [22] developed a computer program for simulation of the dynamic behavior of general
multiple effect evaporation plants. El-Nashar and Qamhiyeh [23] developed a simulation model for predicting the transient
behavior of Multi-effect stack-type (MES) distillation plant. Tonelli et al. [24] presented a computer package for the simulation of MEE for the concentration of liquid foods and studied the dynamic response. Narmine and Marwan [25] developed a
dynamic model for the MEE process to study the transient behavior of the system. Svandova et al. [26] has performed the
steady-state analysis and dynamic simulation of a methyl tertiary-butyl ether (MTBE) reactive distillation. Stefanov and
Hoo [27] developed a distributed parameter model of black liquor falling lm evaporator, based on rst principles knowledge and heat transfer processes for evaporation side of single plate (lamella) of a falling lm evaporator. Stefanov and
Hoo [28] further expanded a single plate model to develop a fundamental model of a falling lm evaporator that accounted
for the condensation side of the plate, the heating/ashing at the evaporator entrance, the evaporator inventory, mixing and
recirculation ows but neglecting the effect of BPR.
In the present investigation dynamic behavior of mixed feed six effect evaporators system is studied. For this purpose the
mathematical model of sextuple effect (tubular falling lm with liquor ow inside the tube) falling lm evaporator system of
a paper industry is developed based on the work reported by Narmine and Marwan [25]. Further the parametric equations
were solved for steady state conditions by vanishing the accumulation terms in model equations by using fsolve solver in
MATLAB source code. To study the transient behavior of the MEE system the solution of simultaneous ordinary differential
equations for sextuple mixed feed evaporator is obtained by using ode45 solver in MATLAB source code and the steady state
solution is used as the initial value.
2. Mathematical modelling
In the present investigation effect of BPR is also considered for and the mathematical modeling is developed for sextuple
falling lm evaporator system with mixed feed based on the work proposed by Narmine and Marwan [25]. In the mixed feed
sequence the feed is given to 5th effect and thick liquor coming out from 5th goes into 6th and from 6th to 4th and from 4th

386

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

Wv3=Wl4-Wl3

Wv2=Wl3-Wl2

Wv4=Wl6-Wl4

Wv5=F-Wl5

Wv6=Wl5-Wl6

Steam

Wl5
Wl3

Wl2

Wl4
Wl6
Feed (F)

Product

c1

c3

c2

c5

c4

c6

Condenser

Fig. 1. Flow diagram of sextuple mixed feed evaporator (? 5 ? 6 ? 4 ? 3 ? 2 ? 1).

Steam/Vapour Inlet

Black
liquor

Steam
Chest

Inlet
Wli+1
Xi+1
Tli+1

Wvi
Tvi

Wvi-1
Tvi-1
hvi-1
ith Effect

Vapour
Outlet

Evaporation
Section

Wli
Xi
Tli

Wvi-1
hci-1

Fig. 2. Block diagram with terms used for the ith effect.

to 3rd and so on up to the rst effect, while the steam input is given in the rst effect as shown in the Fig. 1. The material and
energy ow for ith effect is given in the Fig. 2.
Model equations are developed for ith effect using material and energy balance equations. The equations stating the physical properties of the black liquor are given in Table 1 dened by Gullichsen and Fogelholm [29]. It is assumed that the vapor
generated by the process of concentration of black liquor is saturated. It is also assumed that the energy and mass accumulation in the vapor lump is very small as compared to the enthalpy of the steam and can be neglected.
Material balance for liquor in the ith effect for i = 1, 2 and 3:

d
Mli t Wli1  Wli  Wvi :
dt

Since in the mixed feed sequence the feed is given to fth effect and then thick liquor move from 5th to 6th and from 6th
to 4th effect thus for i = 4, 5 and 6 the equation is slightly different from the backward feed sequence and are given by
Table 1
Physico-thermal properties of black liquor, saturated steam and condensate.
S.
no.

Properties

Equations

1.

Boiling point rise, C

BPR = (6.173X  7.48X1.5 + 32.747X2) {1 + 0.006(Tv  3.7316)}

3

2.

Black liquor density, kg/m

3.

Specic heat capacity, kJ/

kg K
Enthalpy of water, kJ/kg
Enthalpy of saturated
steam, kJ/kg
Latent heat of vaporization,
kJ/kg

4.
5.
6.

Function parameters

Cp = 4.216(1  X) + {1.675 + (3.31 Tl)/1000}X + (4.87  20 Tl/

1000) (1  X)X3
hl = aT + b, where a = 4.1832 and b = 0.127011
hv = c Tv + d, where c = 1.75228 and d = 2503.35

Solid concentration, (X); steam

temperature, (Tv) C
Black liquor temp. (Tl) in C; solid
concentration, (X)
Black liquor temp. (Tl) in C, solid
concentration, (X)
Temp. of water, (T) in C
Steam temperature, (Tv) C

k = 2519.5  2.653Tv

Steam temperature, (Tv) C

Pl = (997 + 649X) {1.008 0.237 (Tl/1000) 1.94(Tl/1000) }

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

For i 4 the material balance equation is

d
Ml4 t Wl6  Wl4  Wv4 ;
dt

Similarly for i 5 the material balance equation is

d
Ml5 t F  Wl5  Wv5 ;
dt

In this way for i 6 the material balance equation is

d
Ml6 t Wl5  Wl6  Wv6 :
dt

387

1a

1b

1c

Energy balance for liquor in the ith effect:

d
Mli thli t Wli1 hli1  Wli hli  Wvi hvi Wvi1 hvi1  W v i1 hci1 ;
dt

where Wvi1 hvi1Qhci1 and also Q Ui Ai Tvi1  T v i  BPRi .

Material balance for solids in ith effect:

d
Mli tXi t Wli1 Xi1  Wli Xi :
dt

The changes shown in material balance for i = 4, 5 and 6 are similarly effective for energy and solid balance and in other
equations which are obtained by using material balances given by (1a), (1b) and (1c).
2.1. Formulation of the system equations
The modeling technique proposed by Narmine and Marwan [25] has been further modied and extended in the present
work to obtain the mathematical modeling of the MEE system. The vapor and liquor in ith effect are in equilibrium and the
relation for the liquor and vapor temperature dened in terms of BPR is as follows:

Tli Tvi BPRi

and the BPR is dened in terms of temperature and solid concentration, presented in Table 1.
Differentiating Eq. (4) with respect to time we get


 




d
d
o
o
o

Tli t
Tvi t 1
BPRi
BPRi
Xi t :
dt
dt
oTv
oX
ot

Mli can be written as:

Mli ALi Pli :

The Pli, the density of the liquor is dened in terms of temperature and solid concentration and also presented in Table 1.
Differentiating Eq. (6) with respect to time we get





d
d
d
Mli t ALi t
Pli APli
Li t :
dt
dt
dt
Since Pli is a function of temperature Tl and concentration X, this equation is reduces into Eq. (7) by using the value of
from Eq. (5) with rearranging the terms.

d
Tli t
dt







d
d
o
o
d
1
Mli t APli
Li t ALi t
Pli
BPRi
Tvi t
dt
dt
oTl
oTv
dt



 

o
o
o
d
Pli
BPRi
Pli
Xi t :
ALi t
oTl
oX
oX
dt

Comparing the resultant differential equation (7) with Eq. (1) we get

C1 C2






d
d
d
Li t C3
Tvi t C4
Xi t ;
dt
dt
dt

where

C1 Wli1  Wli  Wvi ;

C2 APli ;
C3 ALi t

o
Pli
oTl




o
1
;
BPRi
oTv

388

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

Table 2
Range of operational parameters.
S. no.

Operational parameter

Parameters range

Target value

1
2
3
4
5
6
7

Liquor feed ow rate, kg/s

Liquor feed temperature, C
Liquor feed concentration, kg solids/kg total
Steam temperature, C
Last body saturation, C
Liquor product concentration, kg solids/kg total
Heat transfer coefcients (W/m2 K)

18.0025.00
70.0090.00
0.100.52
135140
4852
0.480.52
11601400

23.98
80.00
0.10
139
49
0.50
1160, 1220,1280, 1335, 1365, 1400


C4 ALi t

o
Pli
oTl



 

o
o
:
BPRi
Pli
oX
oX

Differentiating Mli(t)hli(t) with respect to time and using the value of

Eq. (7) we get

d
Tli t
dt

from Eq. (5) and the value of

d
Mli t
dt

from








d
d
d
o
Mli t hli Mli t
hli APli hli
Li t ALi t 1
BPRi
dt
dt
dt
oTv
 
 
 



o
o
d
d
hli
Pli hli
Tvi t ALi t
Xit
Pli
oTl
oTl
dt
dt
 






 
 
o
o
o
o
o
o
hli
BPRi Pli
hli hli
Pli
BPRi
Pli hli :
Pli
oTl
oX
oX
oTl
oX
oX


Comparing the resultant differential equation with Eq. (2) we get


C5 C6






d
d
d
Li t C7
Tvi t C8
Xi t ;
dt
dt
dt

where

C5 Wli1 hli1  Wli hli  Wvi hvi Wvi1 hvi1  W v i1 hci1 ;
C6 APli hli ;
C7 ALi t1

C8 ALi tPli

 



o
o
o
;
BPRi Pli
hli hli
Pli
oTv
oTl
oTl



o
hli
oTl



o
BPEi
oX

o
hli
oX



ALi thli



o
Pli
oTl



o
BPRi
oX

o
Pli
oX


:

d
Differentiating Mli(t)Xi(t) with respect to time and using the value of dt
Mi t from Eq. (7) and rearranging the equation and
representing it in the form of coefcients

Table 3
Steady state solution of sextuple effect evaporator.
Items

Effects
I

II

III

IV

VI

Input liquor concentration (kg/kg)

Output liquor concentration (kg/kg)
Input liquor ow (kg/s)
Output liquor ow (kg/s)
Input liquor temperature (C)
Output liquor temperature (C)
Input steam temperature (C)
Output steam temperature (C)
Input steam ow (kg/s)
Output steam ow (kg/s)
Boiling point rise (C)
Specic heat (kJ/kg K)

0.2883
0.5
8.3173
4.796
99.4662
125.2157
139
111.0319
3.8028
3.5213
14.1838
3.3006

0.2067
0.2883
11.6031
8.3173
83.4384
99.4662
111.0319
94.3051
3.5213
3.2858
5.1611
3.6275

0.1639
0.2067
14.627
11.6031
71.0077
83.4384
94.3051
80.5579
3.2858
3.0239
2.8805
3.7704

0.1401
0.1639
17.1105
14.627
50.419
71.0077
80.5579
69.0649
3.0239
2.4835
1.9428
3.8507

0.1
0.1157
23.98
20.73
80
61.3928
69.0649
60.2439
2.4835
3.25
1.1489
3.9506

0.1157
0.1401
20.73
17.1105
61.3928
50.419
60.2439
49
3.25
3.6195
1.419
3.8924

Heat transfer area (m2)

Stream consumption (kg/s)
Total evaporation (kg/s)
Steam economy

591.9614
3.8028
19.184
5.0447

389

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

th

54

T ( C)

Variation in the Temperature of 6 effect with 10% decrease in feed flow rate
56

52
50
48

3
-3
tx10 (s)

th

Variation in the Temperature of 6 effect with 10% increase in feed flow rate
48

T ( C)

50

46
44
42

3
-3
tx10 (s)

th

Variation in the Product Concentration of 6 effect with 10% decrease in feed flow rate
0.145
X (kg/kg)

0.144
0.143
0.142
0.141
0.140

0.2

0.4

0.6

0.8

1
1.2
-4
tx10 (s)

1.4

1.6

1.8

th

Variation in the Product Concentration of 6 effect with 10% increase in feed flow rate
0.141

X (kg/kg)

0.140
0.139
0.138
0.137
0.136

0.2

0.4

0.6

0.8

1
1.2
-4
tx10 (s)

1.4

1.6

1.8

Fig. 3. Response of 6th effect by disturbing 10% in the feed ow rate.

C9 C10






d
d
d
Li t C11
Tvi t C12
Xi t ;
dt
dt
dt

where

C9 Wli1 Xi1  Wli Xi ;

C10 APli Xi ;
C11 ALi Xi

o
Pli
oTl



o
BPRi
oTv


;




 

o
o
o
:
Pli
BPRi
Pli
C12 ALi Pli Xi
oTl
oX
oX

10

390

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

st

114

T ( C)

Variation in the Temperature of 1 effect with 10% decrease in feed flow rate
115

113
112
111
0

3
-3
tx10 (s)

st

Variation in the Temperature of 1 effect with 10% increase in feed flow rate
112

110

T ( C)

111

109
108
107

1.5
2
2.5
3
-3
tx10 (s)
st
Variation in the Product Concentration of 1 effect with 10% decrease in feed flow rate
0.56
X (kg/kg)

0.5

0.54
0.52
0.50

4
5
-4
tx10 (s)

st

X (kg/kg)

Variation in the Product Concentration of 1 effect with 10% increase in feed flow rate
0.5
0.49
0.48
0.47
0.46

6
-4
tx10 (s)

10

12

Fig. 4. Response of rst effect by disturbing 10% in the feed ow rate.

3. Simulation
3.1. Steady state simulation
To study the dynamic response of any chemical process initial values of the process variables are needed, which can be
obtained using steady state solution. For obtaining the steady state solution of the system, the accumulation terms in the
Eqs. (8)(10) equal to zero i.e. C1 = 0, C5 = 0 and C9 = 0. Further substituting the values of physical properties like enthalpy,
density, specic heat and boiling point rise from Table 1 and heat transfer rate from Eq. (2), mathematical model of 12 nonlinear equations in 12 unknown variables is obtained. The model is similar to the model described by Ray and Sharma [10].
Operational data for simulation is taken from literature. For falling lm evaporator the range of overall heat transfer coefcient (OHTC) is given by various researchers. Gupta [30] has suggested the value of OHTC as 5802900 W/m2 K. Algehed
and Berntsson [31] suggested the values of OHTC nearly 1510 W/m2 K. For evaporation of black liquor Gullichsen and Fogelholm [29] suggested the range of OHTC for all six effects of a sextuple evaporator system between 500 and 2000 W/m2 K. The
values of OHTC depend on the concentration rise in the MEE system. After discussion with few Indian Paper Mills, values of

391

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

th

T ( C)

Variation in the Temperature of 6 effect with 10% decrease in feed concentration

49.15
49.1
49.05
49

3
-3
tx10 (s)

th

49.1

T ( C)

Variation in the Temperature of 6 effect with 10% increase in feed concentration

49.15

49.05
49

3
-3
tx10 (s)

th

X (kg/kg)

Variation in the Product Concentration of 6 effect with 10% decrease in feed concentration
0.145
0.14
0.135
0.13
0.125
0

3
-4
tx10 (s)

th

X (kg/kg)

Variation in the Product Concentration of 6 effect with 10% increase in feed concentration
0.16
0.155
0.15
0.145
0.140
0

3
-4
tx10 (s)

Fig. 5. Response of 6th effect by disturbing 10% in the feed concentration.

OHTC are taken in the range 11601400 W/m2 K for the concentration rise from 10% to 50%. The values of the operating
parameters are presented in Table 2. The steady state solution of the system of nonlinear equations is obtained by fsolve
solver in MATLAB source code and is presented in Table 3. The input values, required steam are shown in bold face and heat
transfer area is calculated and also presented in Table 3. Steam economy dened as the ratio of total evaporation with live
steam consumption is shown in the Table 3. The steam economy for the present system is calculated as 5.0447.
3.2. Dynamic simulation
For the dynamic simulation rst order nonlinear differential Eqs. (8)(10) are solved simultaneously in the same order for
all the six effects of the mixed feed sextuple effect evaporator system. In the mixed feed sequence the feed is given to 5th and
thick liquor coming out from 5th to 6th and from 6th to 4th and so onto the rst effect. This procedure is repeated for each
time steps. Steady-state solution of the model provides initial values of the system variables at time, t = 0. Solution of such
types of simultaneous nonlinear ordinary differential equations is extremely intricate in nature even by using sophisticated
numerical techniques. Stefanov and Hoo [27] solved the system of equations for single plate lamella of a falling lm evaporator by Orthogonal Collocation on nite elements after converting the PDEs into dimensionless form. Stefanov and Hoo
[28] discretized the system of Partial Differential Equations (PDEs) of a MEE system by using Orthogonal Collocation and

392

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

st

111.2

T ( C)

Variation in the Temperature of 1 effect with 10% decrease in feed concentration

111.3

111.1
111
0

3
-3
tx10 (s)

st

111.2

T ( C)

Variation in the Temperature of 1 effect with 10% increase in feed concentration

111.3

111.1
111
0

3
-3
tx10 (s)

st

X (kg/kg)

Variation in the Product Concentration of 1 effect with 10% decrease in feed concentration
0.5
0.48
0.46
0.44

6
-4
tx10 (s)

10

12

st

X (kg/kg)

Variation in the Product Concentration of 1 effect with 10% increase in feed concentration
0.56
0.54
0.52
0.50

6
-4
tx10 (s)

10

12

Fig. 6. Response of rst effect by disturbing 10% in the feed concentration.

then used LSODI (Livermore solver for ordinary differential equations, implicit system) solver, included in ODEPACK given by
Algehed and Berntsson [32] to solve the discretized system of PDEs. The execution time is approximately 14 h on a 1.2 GHz
Athlon node, which is equivalent to 1 h of real time for dynamic simulation of plate type falling lm MEE system for using
backward feed. These techniques are complex and more time consuming. In the present investigation an attempt has been
made for steady state as well as dynamic simulation for tubular type falling lm MEE system for mixed feed by using fsolve
and ode45 solver in MATLAB source code respectively. Simulations results obtained using MATLAB source code shows a
comparable performance with FASTER method in terms of CPU time.

4. Model application
The dynamic response on system variables of MEE is studied by creating four types of disturbances namely (i) in feed ow
rate, (ii) in feed concentration, (iii) in live steam temperature and (iv) in feed temperature as step input function. System
variables selected were (i) temperature of each effect and (ii) output concentration of each effect. The disturbance is taken

393

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

th

Variation in the Temperature of 6 effect with 10% decrease in steam temperature

50

T ( C)

48
46
44
42

3
-3
tx10 (s)

th

Variation in the Temperature of 6 effect with 10% increase in steam temperature

56

T ( C)

54
52
50
48

3
-3
tx10 (s)

th

X (kg/kg)

Variation in the Product Concentration of 6 effect with 10% decrease in steam temperature
0.1402
0.1401
0.14
0.1399
0

3
-4
tx10 (s)

th

X (kg/kg)

Variation in the Product Concentration of 6 effect with 10% increase in steam temperature
0.1406
0.1404
0.1402
0.14

3
-4
tx10 (s)

Fig. 7. Response of sixth effect by disturbing 10% in the steam temperature.

as a step input of 10%. Responses to the disturbance for the last (6th) and rst (rst) effects are presented from Figs. 310.
These graphs are plotted with respect to time to show how the variables approach to the new steady state conditions.
5. Results and discussion
In the mixed feed MEE system such as (6 ? 5 ? 4 ? 3 ? 2 ? 1); the input liquor enters the evaporator system from the
5th effect and then thick liquor from 5th effect to 6th and from 6th to 4th and up to the rst effect. Dynamic responses due
10% step input in last and rst effect in each case of mixed feed are given as follows.
5.1. Effect of varying feed ow rate
In the present case of mixed feed input thick liquor enters the evaporator system from the 5th effect. Thus for the variation
in the feed ow rate the disturbance is applied in the 5th effect. The effect of 10% step input in feed ow rate on the temperature and concentration of last effect (connected with condenser, i.e., 6th) and rst effect (connected with live steam,
i.e., rst) are shown by Figs. 3 and 4, respectively. The temperature and concentration of both the effects show an increase
or decrease with decrease or increase in the feed ow rate. It is obvious as fresh steam supply rate is constant and water
to be evaporates and decreases per unit time. Also the steady state in 6th effect is achieved at a faster rate than the rst effect.

394

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

Variation in the Temperature of 1st effect with 10% decrease in steam temperature

105

T ( C)

110

100
95

0.5

1.5
-3
tx10 (s)

2.5

st

120

T ( C)

Variation in the Temperature of 1 effect with 10% increase in steam temperature

125

115
110
0

3
-3
tx10 (s)

st

Variation in the Product Concentration of 1 effect with 10% decrease in steam temperature

X (kg/kg)

0.502
0.5
0.498
0.496
0.495

4
5
-4
tx10 (s)

st

X (kg/kg)

Variation in the Product Concentration of 1 effect with 10% increase in steam temperature
0.5075
0.506
0.504
0.502
0.5
0.498
0

4
5
-4
tx10 (s)

Fig. 8. Response of rst effect by disturbing 10% in the steam temperature.

5.2. Effect of varying feed concentration

For the dynamic response of the feed concentration the disturbance is applied in the 5th effect. The effect of 10% step
input in feed concentration on the temperature and concentration of last (6th) and rst effect (rst) are shown in Figs. 5
and 6, respectively. The dynamic behavior of effects temperature with respect to disturbances in feed concentration shows
slight but insignicant change in temperature. However the change, it observed is unidirectional i.e. the temperature increases irrespective of increase or decrease in feed concentration. The changes in product concentration of both the effect
show increase or decrease according as the feed concentration is increase or decrease. This may be due to the fact that
DT across the evaporator system remains constant and vapor-liquid equilibrium of each effect remains almost unchanged
for the optimum performance.
5.3. Effect of varying steam temperature
In MEE system with mixed feed, live steam enters in the rst effect. Thus for the variation in the steam temperature the
disturbance is applied in the rst effect. The effect of 10% step input in steam temperature on the temperature and concentration of last and rst effect are shown in Figs. 7 and 8, respectively. The 10% change in steam temperature results in

395

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

th

49

T ( C)

Variation in the Temperature of 6 effect with 10% decrease in feed temperature

49.2

48.8
48.6
48.4

3
-3
tx10 (s)

th

49.6

T ( C)

Variation in the Temperature of 6 effect with 10% increase in feed temperature

49.8

49.4
49.2
49

3
-3
tx10 (s)

Variation in the Product Concentration of 6th effect with 10% decrease in feed temperature

X (kg/kg)

0.1402

0.14015

0.1401
0

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

-4

tx10 (s)
th

X (kg/kg)

Variation in the Product Concentration of 6 effect with 10% increase in feed temperature
0.1402

0.14015

0.1401
0

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

-4

tx10 (s)
Fig. 9. Response of sixth effect by disturbing 10% in the feed temperature.

increase or decrease in the temperature of both the effects before obtaining the steady state for 10% increase or decrease
respectively. 10% disturbance in steam temperature does not result in any noticeable change in the product concentration
of both the effects. However with scale down Y-axis value (up to four decimal point as shown in the plots) it was observed
that product concentration increase and then decrease or decrease and then increase for 10% increase or decrease in the
steam temperature respectively.

5.4. Effect of varying feed temperature

Since feed liquor enters in the 5th effect for the mixed feed, thus the dynamic response of the feed temperature the disturbance is applied in the 5th effect. The effect of 10% step change in feed temperature on the temperature and concentration
of last and rst effect are shown by the Figs. 9 and 10, respectively. It is evident from the gures that 10% disturbance in feed
temperature does not bring noticeable change in the temperature and product concentration each effect. However with scale
down Y-axis (as shown in the plots), it is observed that temperature of the both the effect increases and decreases to obtain the
steady state with an increase and decrease in feed temperature and the product concentration of each effect rst decreases
and then increases to obtain the steady state with a very small uctuations about the steady state up to four to ve decimal

396

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

st

Variation in the Temperature of 1 effect with 10% decrease in feed temperature

111

T ( C)

111.1

110.9
110.8
0

3
-3
tx10 (s)

st

111.6

T ( C)

Variation in the Temperature of 1 effect with 10% increase in feed temperature

111.8

111.4
111.2
111
0

3
-3
tx10 (s)

st

X (kg/kg)

Variation in the Product Concentration of 1 effect with 10% decrease in feed temperature
0.5001
0.5
0.4999
0.4998
0

6
-4
tx10 (s)

10

12

st

X (kg/kg)

Variation in the Product Concentration of 1 effect with 10% increase in feed temperature
0.5002
0.5001
0.5
0.4999
0

6
-4
tx10 (s)

10

12

Fig. 10. Response of rst effect by disturbing 10% in the feed temperature.

places in the value of concentrations of both the effect and conversely for 10% increase in feed temperature, which is negligible
in engineering problems.
6. Conclusions
For the steady state and dynamic simulation the fsolve and ODE45 solvers in MATLAB source code are used successfully.
Simulations results obtained using MATLAB source code show a comparable performance with faster method in terms of CPU
time and conclude that with the use of MATLAB it is possible to have better control of such complex system. Further the
dynamic behavior of each effects temperature and product concentration was studied by disturbing the liquor ow rate,
feed concentration, steam and feed temperatures by 10%. The transient study shows that the steady state is reached more
quickly for temperature in comparison of the solid concentration and all of the responses converge in a smooth fashion.
Acknowledgment
Mr. Deepak Kumar is thankful to Ministry of Human Resource Development; New Delhi, India for providing the Senior
Research Fellowship for this work.

D. Kumar et al. / Applied Mathematical Modelling 37 (2013) 384397

397

References
[1] N.J. Rao, R. Kumar, Energy conservation approaches in a paper mill with special reference to the evaporator plant, in: Proceedings of the IPPTA
International Seminar Conservation in Pulp and, Paper Industry, 1985, pp. 5870.
[2] A.K. Ray, N.J. Rao, M.C. Bansal, B. Mohanty, Design data and correlations of waste liquor/black liquor from pulp mills, IPPTA J. 4 (1992) 121.
[3] W.T. Hanbury, Proc., IDA World Congress on Desalination and Water Sciences. Abu Dhabi, UAE, 4, 1995, p. 375.
[4] H.T. El-Dessouky, I. Alatiqi, S. Bingulac, H.M. Ettouney, Steady-state analysis of the multiple effect evaporation desalination process, Chem. Eng.
Technol. 21 (5) (1998) 437451.
[5] H.T. El-Dessouky, H.M. Ettouney, F. Mandani, Performance of parallel feed multiple effect evaporation system for seawater desalination, Appl. Therm.
Eng. 20 (2000) 16791706.
[6] H.T. El-Dessouky, H.M. Ettouney, F. Al-Juwayhel, Multiple effect evaporation-vapor compression desalination processes, Transactions of IChemE 78
(Part A) (2000) 662676.
[7] A.O.S. Costa, E.L. Enrique, Modeling of an industrial multiple effect evaporator system, in: Proceedings of the 35th Congresso e Exposicao Anual de
Celulose e Papel, 2002.
[8] A.K. Ray, P. Singh, Simulation of multiple effect evaporator for black liquor concentration, IPPTA 12 (3) (2000) 3545.
[9] A.K. Ray, N.K. Sharma, P. Singh, Estimation of energy gains through modelling and simulation of multiple effect evaporator system in a paper mill,
IPPTA 16 (2) (2004) 3545.
[10] A.K. Ray, N.K. Sharma, Simulation of multi-effect evaporator for paper mill-effect of ash and product utilization for mixed feeds sequences, IPPTA 16
(4) (2004) 5564.
[11] V.K. Agarwal, M.S. Alam, S.C. Gupta, Mathematical model for existing multiple effect evaporator systems, Chemical Engineering World 39 (2004) 76
78.
[12] V. Miranda, R. Simpson, Modelling and simulation of an industrial multiple effect evaporator: tomato concentrate, J. Food Eng. 66 (2005) 203210.
[13] M. Karimi, A. Jahanmiri, M. Azarmi, Inferential cascade control of multi-effect falling-lm evaporator, Food Control 18 (2007) 10361042.
[14] D. Kaya, I.H. Sarac, Mathematical modeling of multiple-effect evaporators and energy economy, Energy 32 (2007) 15361542.
[15] R. Bhargava, S. Khanam, B. Mohanty, A.K. Ray, Selection of optimal feed ow sequence for a multiple effect evaporator system, Computers and
Chemical Engineering 32 (2008) 22032216.
[16] R. Bhargava, S. Khanam, B. Mohanty, A.K. Ray, Simulation of at falling lm evaporator system for concentration of black liquor, Computers and
Chemical Engineering 32 (2008) 32133223.
[17] M.H. Khademi, M.R. Rahimpour, A. Jahanmiri, Simulation and optimization of a six-effect evaporator in a desalination process, Chem. Eng. Process. 48
(2009) 339347.
[18] M. Johansson, L. Vamling, L. Olausson, Heat transfer in evaporating black liquor falling lm, Int. J. Heat Mass Transf. 52 (2009) 27592768.
[19] S. Khanam, B. Mohanty, Energy reduction schemes for multiple effect evaporator systems, Appl. Energy 87 (2010) 11021111.
[20] A. Gupta, Mathematical Modeling g and analysis of pulp washing problems, Ph.D. Thesis, Indian Institute of Technology Roorkee, 2001.
[21] H. Andre, R.A. Ritter, Dynamic response of a double effect evaporator, Can. J. Chem. Eng. 46 (1968) 259264.
[22] U. Bolmstedt, Simulation of the steady-state and dynamic behavior of multiple effect evaporator plants, Part: 2 Dynamic Behavior, Comput. Aided Des.
9 (1) (1977) 2940.
[23] A.M. El-Nashar, A. Qamhiyeh, Simulation of the performance of MES evaporators under unsteady state operating conditions, Desalination 79 (1990)
6583.
[24] S. Tonelli, J.A. Romagnoli, J.A. Porras, Computer package for transient analysis of industrial multiple effect evaporators, J. Food Eng. 12 (1990) 267281.
[25] H.A. Narmine, M.A. Marwan, Dynamic response of multi-effect evaporators, Desalination 114 (1997) 189196.
[26] Z. Svandova, M. Kotora, J. Markos, L. Jelemensky, Dynamic behaviour of a CSTR with reactive distillation, Chem. Eng. J. 119 (2006) 113120.
[27] Z.I. Stefanov, K.A. Hoo, Distributed parameter model of black liquor falling-lm evaporators. Part 1. Modeling of a single plate, Ind. Eng. Chem. Res. 42
(2003) 19251937.
[28] Z.I. Stefanov, K.A. Hoo, Distributed parameter model of black liquor falling-lm evaporators. 2. Modeling of a multiple-effect evaporator plant, Ind. Eng.
Chem. Res. 43 (2004) 81178132.
[29] J. Gullichsen, C.J. Fogelholm, Chemical Pulping, Paper Making Science and Technology a series of 19 books, 6B, (2000), B1820.
[30] B.K. Dutta, Heat Transfer Principles and Applications, Prentice Hall of India Pvt. Ltd, New Delhi, 2005.
[31] J. Algehed, T. Berntsson, Evaporation of black liquor and wastewater using medium-pressure steam: Simulation and economic evaluation of novel
designs, Appl. Therm. Eng. 23 (2003) 481495.
[32] A.C. Hindmarsh, ODEPACK a systemized collection of ODE solvers. Scientic Computing; 1 of IMACS Transactions on Scientic Computing; IMACS,
1982, 5564.